多向度電腦化適性測驗在發展遲緩兒童篩檢之應用

Application of Multi-dimensional Computerized Adaptive Test in the Screening for Developmental Delay Children

探討利用多向度電腦化適性測驗（NICAT評估發展遲緩兒童之可行性。利用文獻發展的5構面（即認知、語言、大動作，細動作、及適應社會）75道題庫的發展遲緩試題參數，模擬常態分佈1000受試者的MCAT發展遲緩兒童回答各題的反應分數（0及1），與全部作答（NAT的精準度不相等分數比率小於5%）及效率性（答題數量節省1／3以上），及決定其決斷點。驗證MCAT較傳統全部作答來得更有效率（題目數量節省46.67%=（75－40）75）。模擬1000人的MICAT各構面高度相關。受試成績不同百分比小於5%的t檢定顯示能力估計值沒有顯著差異。總體決斷點分數可以設定於－－0.7與0.7loit，即33%及67%。利用MCAT評估發展遲緩兒童之可行性會比NAT更有效率，卻不失其測量的精準度。發展網路平台，方便家長使用電腦、平板或手機自行在線上進行發展遲緩兒童之評估，將有助於篩檢之推廣。 Using multidomain developmental screening tools is a feasible method for pediatric healthcare professionals to identify children at risk for developmental problems. However, the applicability of a specialized multidimensional computer adaptive testing (MCAT) tool is still unknown. We investigated using MCAT combined with Multidimensional Screening in Child Development (MuSiC) for toddlers. After we had retrieved the parameters for 75 items from the MuSiC literature item bank for 1- to 3-year-olds, we simulated 1,000 person measures from a normal standard distribution to compare the efficiency and precision of MCAT and NAT (Non-Adaptive Testing) in five domains: cognitive skills, language skills, gross motor skills, fine motor skills, and socio adaptive skills. The number of items saved and the cutoff points determined for the tool were determined. MCAT yielded significantly more precise measurements and was significantly more efficient than was NAT: it yielded a 46.67% (= 75-40)/75) saving in item length when measurement differences less than 5% were allowed. Person-measure correlation coefficients were highly consistent among the five domains. The cutoff points for the overall measures were -0.7 and 0.7 logits, which was equivalent to 33 and 67 in percentile scores. Significantly fewer items were answered on MCAT than on NAT without compromising MCAT’s precision. Developing a website to help parents with their own computers, tablets, or smart phones for online screening and prediction of developmental delays in toddlers is useful and not difficult.

Using multidomain developmental screening tools is a feasible method for pediatric healthcare professionals to identify children at risk for developmental problems. However, the applicability of a specialized multidimensional computer adaptive testing (MCAT) tool is still unknown. We investigated using MCAT combined with Multidimensional Screening in Child Development (MuSiC) for toddlers. After we had retrieved the parameters for 75 items from the MuSiC literature item bank for 1- to 3-year-olds, we simulated 1,000 person measures from a normal standard distribution to compare the efficiency and precision of MCAT and NAT (Non-Adaptive Testing) in five domains: cognitive skills, language skills, gross motor skills, fine motor skills, and socio adaptive skills. The number of items saved and the cutoff points determined for the tool were determined. MCAT yielded significantly more precise measurements and was significantly more efficient than was NAT: it yielded a 46.67% (= 75-40)/75) saving in item length when measurement differences less than 5% were allowed. Person-measure correlation coefficients were highly consistent among the five domains. The cutoff points for the overall measures were -0.7 and 0.7 logits, which was equivalent to 33 and 67 in percentile scores. Significantly fewer items were answered on MCAT than on NAT without compromising MCAT’s precision. Developing a website to help parents with their own computers, tablets, or smart phones for online screening and prediction of developmental delays in toddlers is useful and not difficult.